Trusting Your Evidence: Hallucinate Less with Context-aware Decoding

Code

• github.com/xhan77/context-aware-decoding
  pytorch★ 60
• github.com/hongshi97/cad
  none★ 33
• github.com/danshi777/ircan
  pytorch★ 12

Abstract

Language models (LMs) often struggle to pay enough attention to the input context, and generate texts that are unfaithful or contain hallucinations. To mitigate this issue, we present context-aware decoding (CAD), which follows a contrastive output distribution that amplifies the difference between the output probabilities when a model is used with and without context. Our experiments show that CAD, without additional training, significantly improves the faithfulness of different LM families, including OPT, GPT, LLaMA and FLAN-T5 for summarization tasks (e.g., 14.3% gain for LLaMA in factuality metrics). Furthermore, CAD is particularly effective in overriding a model's prior knowledge when it contradicts the provided context, leading to substantial improvements in tasks where resolving the knowledge conflict is essential.

Reproductions